It has been common to have elevator mechanics check brake operation visually by determining when the actual braking operation begins by visual measuring of distance. Such a test is subject to human error: for example, an error of only 100 microseconds in determining the actual beginning of the braking operation will result in an error of one quarter of a meter if the elevator speed were 2.5 meters per second. In certain modern elevators operating at 10 meters per second, the error would be a full meter. Such tests also require that the elevator be taken out of service for some period of time. The test can only be performed with a mechanic at the elevator site, and will require between five minutes and twenty minutes of the mechanic's time to carry out the test. Such tests are only qualitative, resulting in pass/fail or poor/fair/good indications of results.
More recently, external devices have been utilized to measure the parameters of the elevator brake system. Such devices are usually quite complex, require additional hardware attached to the elevator, are difficult to operate, and require great expertise in order to interpret the result.
Any of the tests overtly performed with human intervention must be performed according to a schedule, such as at regular intervals of time, or a schedule based upon elevator usage.